Pancreatic cancer is an extremely lethal disease with the lowest 1-year and 5-year survival rates of any cancer. This is due, in part, to the extremely metastatic behavior of pancreas carcinoma cells, which are also highly resistant to therapy. Importantly, we now know that a strong, but nevertheless unique, stromal response is present in pancreatic ductal adenocarcinoma (PDA). This is highly relevant as it is now recognized that, in many solid tumors, the local microenvironment and the stromal compartment significantly influences disease progression. In conventional pancreatic ductal adenocarcinoma, and its most common precursor PanIN lesions, disease progression is associated with a robust fibrotic response in the stroma, or desmoplastic reaction, that is largely regulated by pancreatic stellate cells. Even at the early stages of preinvasive disease activation of pancreatic stellate cells and extracellular matrix deposition is robust. Through disease progression this desmoplastic reaction continues and often intensifies, offering critical support to carcinoma cells as they progress to fully metastatic disease while also providing drug-free sanctuaries that limit access of chemotherapeutic agents. However, to date, the molecular and physical mechanisms by which PSCs regulate epithelial carcinoma cell behavior in vivo are not well understood. Likewise, the molecular mechanisms by which PSCs are activated as well as the source of activated PSCs in the tumor remains to be fully elucidated. In culture, stellate cells are known to secrete factors that can promote cell behaviors associated with tumor progression, suggesting a paracrine signaling role for stellate cells in indigenous disease in vivo. Therefore, here, using murine models of pancreas cancer that faithfully mimics the human disease and human cell line grafted tumors, we propose specific experiments to explicitly investigate the source of activated pancreatic stellate cells in pancreas cancer, dissect the role of stromal stellate cells population in preinvasive, invasive and metastatic disease, and specifically target a key stellate cell-derive cytokine that promotes carcinoma cell growth and motility and contributes to the generation and maintenance of the desmoplastic reaction in PDA. We hypothesize that bone marrow-derived progenitor cells are a source of activated pancreatic stellate cells that are co-opted, along with endogenous populations of stellate cells, in early preinvasive lesions to help drive conversion to PDA and ultimately establish successful metastases. Furthermore, we hypothesize that chemical signal draws bone marrow-derived progenitor cells to PDA but that physical features of the stroma also influence their infiltration and subsequent differentiation and that specific strom targeting therapy can disrupt these interactions and the tumor supporting influence of stellate cells in PDA.